Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

1. Can you support pytorch<2.0 for training? Computers do not support cuda for 11.8 #2

Open
Kidgod77 opened this issue May 4, 2023 · 1 comment
Labels
wontfix This will not be worked on

Comments

@Kidgod77
Copy link

Kidgod77 commented May 4, 2023

  1. Can you support pytorch<2.0 for training? Computers do not support cuda for 11.8
    2 What are the differences between train experiment files and train? Is there a problem with the train file code? train experiment file I can use cpu to train, train will report error.
@Daraan
Copy link
Owner

Daraan commented May 4, 2023

I'll have a cleaning of the files & readme planned. Was quick release. Not sure if I get to it within this month.

  1. PyTorch < 2.0 is in general possible and the first versions worked on it. Currently cannot estimate what needs to be changed to make the versions compatible again and what might break. I mostly worked with cuda 11.7.

  2. The train.py file is set up to work with a single settings.json file and was build more with DeepSpeed in mind, the train_experiments.py to works with several input files from the experiments folder and is more recent.
    I'm planning to unify these.

train.py file has been fixed thank you. The --logname parameter was not optional like intended.

@Daraan Daraan added the wontfix This will not be worked on label May 23, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
wontfix This will not be worked on
Projects
None yet
Development

No branches or pull requests

2 participants